FACTORS INFLUENCING THE USAGE BEHAVIOR OF LARGE LANGUAGE MODELS TO SUPPORT LEARNING AMONG UNDERGRADUATE STUDENTS IN BANGKOK, THAILAND

Abstract

This study investigates the causal factors that influence university students’ intention to use, and actual usage of, large language models (LLMs) as learning support tools in Bangkok, Thailand. The research integrates three prominent frameworks—Technology Acceptance Model, Theory of Planned Behavior, and Unified Theory of Acceptance and Use of Technology—into a single comprehensive model. Data were collected from 400 undergraduate students  in Bangkok Thailand selected through stratified random sampling. A structured questionnaire using Likert-type rating scales was administered, and the data were analyzed using descriptive statistics, multiple regression analysis, and analysis of variance. The findings indicate that perceived usefulness, perceived ease of use, social influence, perceived behavioral control, and hedonic motivation have significant positive effects on behavioral intention to use LLMs, with perceived usefulness emerging as the strongest predictor. Actual usage behavior is mainly determined by facilitating conditions and habit, with habit exerting the greatest predictive power. Ethical awareness functions as a negative moderating variable that weakens the relationship between perceived usefulness and behavioral intention, whereas artificial intelligence knowledge serves as a positive moderator that strengthens the relationship between perceived ease of use and behavioral intention. The results also show significant differences across years of study, as third- and fourth-year students report more intensive LLM usage than first- and second-year students. Overall, the study suggests that LLMs have become embedded in undergraduate learning routines. It recommends that higher education institutions develop policies, curricula, and support systems that foster digital intelligence, improve facilitating conditions, and provide guidance for responsible and ethical LLM use to protect academic integrity while maximizing educational benefits.